baikal.Model.fit

Model.fit(X, y=None, **fit_params)

Trains the model on the given input and target data.

The model will automatically propagate the data through the pipeline and fit any internal steps that require training.

Parameters
  • X

    Input data (independent variables). It can be either of:

    • A single array-like object (in the case of a single input)

    • A list of array-like objects (in the case of multiple inputs)

    • A dictionary mapping DataPlaceholders (or their names) to array-like objects. The keys must be among the inputs passed at instantiation.

  • y

    Target data (dependent variables) (optional). It can be either of:

    • None (in the case all steps are either non-trainable and/or unsupervised learning steps)

    • A single array-like object (in the case of a single target)

    • A list of array-like objects(in the case of multiple targets)

    • A dictionary mapping target DataPlaceholders (or their names) to array-like objects. The keys must be among the targets passed at instantiation.

    Targets required by steps that were set as non-trainable might be omitted.

  • fit_params – Parameters passed to the fit method of each model step, where each parameter name has the form <step-name>__<parameter-name>.